Nine separate projects related to outcomes of patients with rheumatic diseases are included in this report. The first project, Genetic determinants of ankylosing spondylitis severity, is a prospective observational study of subjects with ankylosing spondylitis (AS) that seeks to identify genetic determinants of AS susceptibility and severity. This project will test genotype-phenotype correlations in a large sample. Over 1200 subjects have been enrolled at five centers including the NIH Clinical Center, and genetic testing and data analysis are underway. Previous findings include identification of 26 new genes associated with susceptibility to AS, gene-gene interaction between HLA-B27 and ERAP1 in risk of AS, lack of association of bone metabolism candidate genes with radiographic severity, impact of hip arthritis on physical functioning, lessening of association between radiographic damage and functional impairment with more prolonged AS, and association of ESR and CRP with axial impairments. Current work examines the association between treatment with tumor necrosis factor alpha inhibitors and changes in the progression of radiographic spine fusion. Collaborators include Drs. J. Reveille, M. Brown, M. Weisman, L. Gensler, and T. Learch. Ongoing work will test fine mapping and additional susceptibility markers in larger samples. The second project, Progression of spinal fusion in ankylosing spondylitis, is a developmental study to test a measure of spinal fusion in AS based on quantification of syndesmophytes in the intervertebral discs by computed tomography. Fifty-nine subjects have been enrolled. Computerized semi-automatic algorithms for measuring syndesmophyte volume and height have been optimized to maximize reliability. Based on the three-dimensional information provided by these scans, we have discovered that syndesmophytes are preferentially formed at the posteriolateral vertebral rim. This localization coincides with areas of high mechanical stress and suggests that biomechanics are likely under-recognized factors in syndesmophyte development. We have tested this hypothesis using finite element analysis on longitudinal data. We have recently extended the imaging to include the thoracic spine, and have found that the thoracic spine is more commonly affected by syndesmophytes than the lumbar spine, and changes in these regions are not correlated. We are currently examining the dynamics of syndesmophyte growth within different regions of the disc space, and planning a study of CT imaging using reduced doses of radiation. Collaborators on this project are Drs. J. Flynn, L. Yao, Y. Yao, and S. Tan. The third project, Clinically important changes in rheumatoid arthritis, is a prospective observational study of clinically important changes in rheumatoid arthritis (RA) activity. Based on longitudinal data on 250 patients, criteria for improvement have been determined for pain, physical functioning, patient global assessment, and four composite measures of RA activity, as well as for SF-36 scales. We also determined that clinical trial response criteria, such as the ACR20, are sensitive but not specific measures of improvement. We have also determined that correlates of the patient global assessment differ with the level of RA activity, and that patient-physician discrepancies are due in part to use of different standards of comparison. We have used these data to validate the health transition question, and found that generic transition questions are as valid as domain-specific transition questions. We have recently developed statistical equivalences between changes in composite RA activity measures, such as the DAS28 and SDAI, and changes in ACR20 and ACR50 responses for use in the planning and interpretation of clinical trials. The fourth project, Outcomes in patients with RA, uses secondary data to examine risks of mortality, malignancy and cardiovascular disease between patients with RA and those without RA. The goals of the fifth project, Clinical epidemiology of systemic lupus erythematosus, are to investigate health disparities among patients with SLE, and to identify clinical features and health care practices that are associated with health outcomes. We have completed a systematic literature review and Bayesian meta-analysis of end-stage renal disease risk in patients with lupus nephritis, which document improved outcomes between 1970 and 1995, but no subsequent improvement in risk of end-stage renal disease, along with a slight increase recently. We recently used the same methodological approach to study trends in mortality among patients with SLE over time. Our results indicate that there has been no improvement in survival in patients with SLE since the early 1990s. We have also recently determined that, among patients with end-stage renal disease, the survival advantage of blacks on dialysis relative to whites on dialysis is due to differential use of transplantation and withdrawal of dialysis. We are currently examining inter-hospital differences in mortality among hospitalized patients with SLE. The goal of the sixth project, Outcomes in Orthopedics, is to investigate associations between processes and outcomes of orthopedic care. Risk of atypical hip fractures was associated with degree of compliance with bisphosphonate treatment among Medicare beneficiaries. In an observational study we found no difference in subtrochanteric beaking and chronic bisphosphonate use, indicating that this bone change is not an inevitable consequence of bisphosphnate treatment. We also reported an association between media reports of the association between bisphosphonate use and atypical fractures and a subsequent decline in these fractures. We also reported than only one-half of Medicare beneficiaries with osteoporosis were treated with anti-resorptive medications. The seventh project examines the frequency and complications of orthopedic procedures among Medicare beneficiaries. We have found that initial treatment of hip fracture with hemiarthroplasty is equally effective as total hip replacement. We have found that patients with ankylosing spondylitis not only have higher rates of total hip arthroplasty than patients without ankylosing spondylitis, but also have higher rates of total knee arthroplasty. Risk of knee arthroplasty was higher among patients who also had hip arthroplasty, suggesting that altered biomechanics may have a role in knee damage. We are currently analyzing associations between treatment with tumor necrosis factor alpha inhibitors and risk of total hip arthroplasty in patients with ankylosing spondylitis. The goal of the eight project is to test whether patients with RA in clinical remission can safely be withdrawn from treatment with TNF inhibitors without relapse of their arthritis. We have initiated a multicenter double-blind placebo-controlled withdrawal trial to test this hypothesis in 300 patients with RA in remission. This study will also provide data on clinical, imaging (joint ultrasound and magnetic resonance imaging), and immunological predictors of relapse. Patient enrollment is underway; 62 patients have been randomized to date. The ninth project is a study of the etiology of melorheostosis, a progressive bone-forming disease, which is testing the hypothesis that somatic mutations in affected bone are responsible for the condition. Bone biopsies and extensive clinical and radiographic assessments are underway. Initial genetic studies of affected and unaffected bone support the presence of somatic mutation as causative in one-half of patients.